All retroviruses generate a full-length 7-9 kb primary RNA transcript, which is the unspliced mRNA for Gag and Gag-Pol polyproteins and the genomic RNA that is packaged into viral particles. This unspliced RNA is normally quite stable. However, a premature termination codon in the gag gene triggers nonsense-mediated mRNA decay (NMD) that rapidly degrades the RNA upon translation. This old result has become surprising in light of current popular models that propose that a downstream intronic sequence, marked during splicing by an exon-junction complex, is necessary to distinguish premature from natural termination codons in higher eukaryotes. In addition, Rous sarcoma virus, a simple avian retrovirus, has a 400 nt cis-acting stability element, which is located immediately downstream of the gag termination codon. Deletions in this 3' UTR region cause the RNA to undergo NMD, forcing the natural gag termination codon to be recognized as a premature termination codon by the cellular machinery. This suggests that the downstream context is important in determining proper translation termination and RNA stability. The stability element will also stabilize a PTC in gag, when inserted directly downstream of it. This stability element may be especially important for retroviral RNAs, since they have an unusually long 3' untranslated region, which usually destabilizes mRNA. In this proposal, both the mechanism of NMD and of immunity to NMD in retroviral RNA will be investigated. Differences in RNA structure or RNP composition are hypothesized to distinguish the coding region (faux UTR) from the 3' UTR downstream from the natural termination codon. Initial studies will be carried out with RSV but extended to a number of other retroviruses, including HIV-1. Intracellular sites of viral RNA decay will also be investigated.